Warning: unmigrated tfhub.dev model artifacts will be deleted on March 18, 2024.

As of November 15th 2023, most tfhub.dev URLs and model handles are now redirecting to their migrated/equivalent counterpart on Kaggle Models.

On March 18, 2024, all unmigrated model assets previously surfaced on tfhub.dev will be deleted – after this date, hub.load and hub.KerasLayer calls to these tfhub.dev handles will fail permanently. See the list of unmigrated model assets here:

• inaturalist/vision/embedder/inaturalist_V2
• nvidia/unet/industrial/class_1
• nvidia/unet/industrial/class_2
• nvidia/unet/industrial/class_3
• nvidia/unet/industrial/class_4
• nvidia/unet/industrial/class_5
• nvidia/unet/industrial/class_6
• nvidia/unet/industrial/class_7
• nvidia/unet/industrial/class_8
• nvidia/unet/industrial/class_9
• nvidia/unet/industrial/class_10
• silero/silero-stt/de
• silero/silero-stt/en
• silero/silero-stt/es
• svampeatlas/vision/classifier/fungi_mobile_V1
• svampeatlas/vision/embedder/fungi_V2

If you are an owner of an unmigrated model, please get in touch with us at [email protected] if you'd like to migrate your model. If you take no action, your model(s) will be deleted on March 18, 2024 and not retrievable (either by you or other users).

For models with a Kaggle Models copy, there will be no impact on the availability/functionality of models that were copied from tfhub.dev – tensorflow_hub will continue to support downloading models that were initially uploaded to tfhub.dev via e.g. hub.load("https://tfhub.dev/<publisher>/<model>/<version>"). To see if a tfhub.dev model has been migrated, enter the model handle in your URL bar – if the redirect is successful, it has already been migrated, otherwise it is an unmigrated model and will be subject to deletion.

Although no migration or code rewrites are explicitly required, we recommend replacing tfhub.dev links with their Kaggle Models counterparts to improve code health and debuggability.

See FAQs here.

This GitHub repository hosts the tensorflow_hub Python library to download and reuse SavedModels in your TensorFlow program with a minimum amount of code, as well as other associated code and documentation.

• Introduction
• The asset types of tfhub.dev
  • SavedModels for TensorFlow 2 and the Reusable SavedModel interface.
  • Deprecated: Models in TF1 Hub format and their Common Signatures collection.
• Using the library
  • Installation
  • Caching model downloads
  • Migration to TF2
  • Model compatibility for TF1/TF2
  • Common issues
  • Build from source
  • Hosting a module
• Tutorials
  • TF2 Image Retraining
  • TF2 Text Classification
  • Additional TF1 and TF2 examples

If you'd like to contribute to TensorFlow Hub, be sure to review the contribution guidelines. To contribute code to the library itself (not examples), you will probably need to build from source.

This project adheres to TensorFlow's code of conduct. By participating, you are expected to uphold this code.

We use GitHub issues for tracking requests and bugs.

Apache License 2.0